Deskripsi Pekerjaan
At Nexus AI Labs, we are building the next generation of generative AI infrastructure. We are looking for a Senior Machine Learning Engineer to lead the design, training, and deployment of large-scale foundation models. You will work at the intersection of research and production, optimizing distributed training pipelines and pushing the boundaries of model efficiency.
This is a high-impact role where your work directly shapes our core product. You will collaborate with a world-class team of researchers and engineers in a fast-paced, innovation-driven environment. We value deep technical expertise, ownership, and a passion for solving hard ML systems problems.
Why join us?
- Work on cutting-edge LLM/GenAI architecture at scale.
- Competitive equity package with low strike price.
- Top-tier hardware budget (H100/A100 clusters).
- Comprehensive health, dental, vision, and 401k matching.
- Hybrid flexibility: 3 days onsite in our SoMa office.
Tanggung Jawab
- Design and implement distributed training frameworks for billion-parameter models using PyTorch, FSDP, and DeepSpeed.
- Optimize model inference latency and throughput via quantization (GPTQ/AWQ), distillation, and custom CUDA kernels.
- Architect and maintain robust MLOps pipelines (data versioning, experiment tracking, CI/CD for ML) using Kubernetes and Kubeflow.
- Collaborate with Research Scientists to translate novel architectures into production-grade code.
- Conduct rigorous ablation studies and benchmarking to drive model performance improvements.
- Mentor junior engineers and establish best practices for code quality, testing, and documentation.
- Participate in on-call rotation for critical model serving infrastructure.
Kualifikasi
- M.S. or Ph.D. in Computer Science, Machine Learning, or related quantitative field (or equivalent industry experience).
- 5+ years of professional experience building and deploying deep learning systems in production.
- Expert proficiency in Python and PyTorch; strong C++/CUDA experience for kernel optimization.
- Proven track record of training large models on multi-node GPU clusters (100+ GPUs).
- Deep understanding of Transformer architecture, attention mechanisms, and scaling laws.
- Extensive experience with distributed computing paradigms (DDP, FSDP, Megatron-LM, Ray).
- Strong software engineering fundamentals: Git, Docker, Kubernetes, CI/CD, testing.
- Excellent communication skills; ability to articulate complex technical concepts to cross-functional stakeholders.